US, China Plan First Official AI Safety Talks Ahead of Trump-Xi Summit

Artificial intelligence has become a central fault line in the US-China technology rivalry, yet the two governments lack a regular channel for managing risks from increasingly capable models. The proposed meeting would mark their first official bilateral dialogue focused on AI safety, including the threat of AI-enabled cyberattacks and the role of voluntary safeguards and self-regulation at leading laboratories.

The two sides are preparing to hold the talks in mid-September, ahead of a planned Trump-Xi summit in Washington later that month, according to the report. US officials are also expected to raise concerns about proprietary American AI models being replicated through model distillation. The participants and exact date have not been disclosed, indicating that arrangements remain under negotiation.